This revised proposal has as an objective the development of a novel therapy for well-differentiated oral squamous carcinoma based on targeting the expression of an endoplasmic reticulum resident protein that is uniquely expressed on the cell surface of well-differentiated carcinomas. Well-differentiated oral squamous cell carcinomas like other well-differentiated carcinomas are difficult to treat because of their low mitotic indices and proliferation rates. However, these well- differentiated neoplasms are known to possess and express an endoplasmic reticulum (ER) collagen chaperone, Hsp47. Although Hsp47 specifically binds procollagen in the ER, in malignancy the protein escapes ER retention to be expressed on the cell surface. We hypothesize that the specific peptide binding characteristics of this protein and unique location in malignancy provides a marker that may serve as a target to which drugs or contrast agents may be directed for chemotherapy or imaging. This hypothesis will be tested through the accomplishment of three specific aims. These include: (1) Expand the repertoire of Hsp47-binding peptides by utilizing random peptide libraries and two-hybrid screening of random displayed peptides with Hsp47 as a bait protein; (2) Determine the availability and fate of Hsp47 binding peptides on the cell surface of epidermoid carcinoma cells in culture and in solid tumors; and (3). Determine the efficacy of Hsp47-binding peptides and Hsp47 monoclonal antibodies in homing chemotherapeutic drugs to tumor sites in oral squamous carcinoma xenografts. To accomplish these aims we have assembled a collaborative team of pathologists, molecular biologists, oncologists, and experts in developmental therapeutics that encompass the University of Maryland's Schools of Dentistry, Medicine and The Greenbaum Cancer Center. The ultimate success of this proposal will be determined by the impact that such an approach has on reducing the morbidity and mortality of oral cancer.